Constraint-Driven Instructions Selection and Application Scheduling in the DURASE system

Kevin Martin (1) , Christophe Wolinski (1) , Krzysztof Kuchcinski (2) , Antoine Floch (1) , François Charot (1)

1 CAIRN - Energy Efficient Computing ArchItectures with Embedded Reconfigurable Resources
2 Department of Computer Science [Lund]

This paper presents a new constraint-driven method for computational pattern selection, mapping and application scheduling using reconfigurable processor extensions. The presented method is a part of DURASE system (Generic Environment for Design and Utilization of Reconfigurable Application-Specific Processors Extensions). The selected processor extensions are implemented as specialized processor instructions. They correspond to computational patterns identified as most frequently occurring or other interesting patterns in the application graph. Our methods can handle both time-constrained and resource-constrained scheduling. Experimental results obtained for the MediaBench and MiBench benchmarks show that the presented method ensures high speed-ups in application execution.

Mots clés: ASIP; reconfigurable computing; scheduling; instruction selection; constraint programming

Domaines: Architectures Matérielles [cs.AR]

Kevin Martin, Christophe Wolinski, Krzysztof Kuchcinski, Antoine Floch, François Charot. Constraint-Driven Instructions Selection and Application Scheduling in the DURASE system. 20th IEEE International Conference on Application-specific Systems, Architectures and Processors, (ASAP 2009), Jul 2009, Boston, United States. pp.145-152, ⟨10.1109/ASAP.2009.19⟩. ⟨inria-00449747⟩